Prostate cancer (PCa) is one of the most frequently diagnosed cancers in men with the average age of diagnosis at 66 years. Accuracy in the early characterization of PCa is the major unmet need in the disease management in order to stratify patients with indolent disease or patients with high risk for aggressive disease at a very early stage. According to guidelines of the world health organization, patient management relies on the Gleason score. A lower-grade cancer usually grows slowly and is less likely to spread than a high-grade cancer, which usually requires aggressive therapy, and in some cases even prostatectomy. A Gleason score of 7 indicates a medium-grade cancer for which the clinical management is hard and subjective, unavoidably leading many patients to either over or under-treatment with devastating consequences. This project’s aim is to provide computer assisted diagnosis regarding the gray zone of Gleason 7 prostate cancer characterization towards eliminating human-judgement related errors based on objective evidence from both MRI and US-guided biopsy. To this end, a multi parametric imaging protocol and multiple location biopsy samples coupled with a machine-learning framework is proposed for the accurate staging and consequently the definition of the most effective treatment plan.
Patients with Gleason score around 7 run a risk of being over or under-treated due to the lack of precision diagnosis in current clinical practices. This project innovates in a twofold manner: firstly, by embracing prostate tissue heterogeneity by collecting and individually pathologically assessing the Gleason score in 12 tissue biopsy samples per patient accurately aligned to their MR imaging position (SO1) and, secondly, by implementing a novel radiomics-pathomics diagnostic workflow (SO2). Critically the proposed framework will offer diagnostic support prior to the decision of aggressively treating the patient by radical prostatectomy with the ultimate goal to spare patients from unnecessary, treatment with devastating consequences in their quality of life. This project aims at building a complete subjective diagnostic computational tool that will assist in the objective classification of patients above or below the critical threshold of Gleason 7 from the early stage of imaging, as these patient cohorts need to be enrolled for very different therapeutic schemas. Gleason >7 entails high risk for metastasis while Gleason < 7 is more conservatively treated. It is of utmost importance to objectively assess all available information reducing the sensitivity related to human nature and totally relying on evidence-based findings. In conclusion, there are no previous efforts to combine MRI radiomics and local pathomics from corresponding biopsy samples to compose a descriptive signature of tissue characteristics from extensive tissue sampling, used for evidence-based in-vivo risk assessment of PCa.
